A Modified Binary PSO Algorithm for Scheduling Independent Jobs in Grid Computing System

Abstract

Grid computing has been considered as an efficient high performance computing platform to solve large scale and complex computing problems. In computational Grids, the Grid scheduler schedules the submitted jobs and finds the appropriate available resource for each job. Job scheduling which is one of the NP-complete problems has been a focus of many researchers in Grid computing area. This paper presents a modified binary particle swarm optimization (MBPSO) algorithm for efficiently allocating jobs to resources in a Grid system so that makespan and flowtime are minimized. The extensive experimental study shows that our modified binary PSO based scheduler outperforms classical PSO based scheduler and also reveals its efficiency when makespan and flowtime are minimized in varied circumstances.